U N I V E R S I T Y OF THE WITWATERSRAND

A F R I C A N S T U D I E S I N S T I T U T E

African Studies Seminar Paperto be presented in RW4.00pro SEPTEMBER 1988

Title: The Mfecane: Beginning the Inquest.

by: John Wright and Julian Cobbing

No. 241A

JETTISONING THE MFECANE (WITH PERESTROIKA)*

Julian Cobbing/August 1988

In this paper I elaborate on the argument that 'the mfecane' is apivotal component of a 'liberal1, settler, apartheid-skeletal form a newanalysis- The main assertion of mfecane propaganda is that a'Zulu-centric' revolution produced an extensive depopulation whichexplains in historiographical sequence: the f1ight of peoples into the'liberation' of the European economy, the land division of 1913, and,since the 1950s, the configuration of the Bantustans. In reply, it isshown that the sub-continental destabilisations and transformationswithin black societies sprang from the synchronous and converging impactof European penetration at Delagoa Bay, the Cape, north of the Orange,and Natal. In order to disguise what had occurred the whites erasedthemselves from their own impact, and retrospectively inserted Shaka andother victims of the process as initiators in situations where they wereabsent. The chronology is lengthened far beyond the (in this context)irrelevant reign of the Zulu monarch. Particular attention is paid tothe sequences of this extended chronology and to the cross-interactionsbetween the sectors of the white advance. It is not the intention tominimise change internal to black societies, but rather to make a callfor this to be researched in its proper context. The huge gaps in ourknowledge revealed by this approach ensure that this task is aformidable one.

Where we are and how we got there

The basic propositions of mfecane propaganda are blissfully simple (anecessary attribute of myth: see Barthes). The 1980s version - aproduct of the refinements of Oner-Cooper and The Oxford^ History - hasan 'explosion' amongst the northern 'Nguni1 triggered by overpopulationwhich (somehow) led to the hegemony of the Zulu 'empire' of Shaka. Thisuniquely revolutionary and predatory state depopulated Natal and forcedneighbours in flight into the interior where they set up 'shock waves'(the semiology of mfecane literature would repay a study) over half ofAfrica. The instantly hZulu-ised' Ndebele depopulated the Transvaalbefore being chased (by the Zulu) into Zimbabwe. Fleeing Dlamini groupsformed Swaziland in the eastern escarpment. Gaza and Jere refugees fromShaka devastated the Delagoa Bay area (or perhaps the Zulu did), theJere, i.e. the Ngoni, creating a Zulu-inspired havoc as far as LakeNyanza. The equally Zulu-ised Ngwane of Matiwane marauded into theCaledon and 'set into motion1 the Tlokwa, that is Mantatees of MaNtatisiwho, briefly Zulu-ised herself (though only for three years) depopulatedthe Orange Free State, and - either the Tlokwa or tertiary victims in

* The arguments in this paper are not necessarily those of John Wright.It is a provisional study for discussion: if you wish to quote from itplease obtain clearance from me at Tel (0461) 26365/22023. Abibliographical note will be available at the seminar.

the chain reaction such as Sebetwane's Kololo: there is no unanimity -were only thrown back from an attempted invasion of the Cape Colony atthe heroic battle of Dithakong in 1823. In the Caledon Moshoeshoegathered exhausted survivors and began to form Basutoland, an island ofsecurity in the Zulu-inspire holocaust. The Ngwane meanwhile metnemesis when their next proposed victims, the Tembu, called in a Britishcommando to their rescue (Mbolompo: 1828). Survivors of perhaps 720,000(sic) peoples fleeing Shaka through and out of Natal found only briefrespite amongst the Gcaleka of Hintsa before succumbing to newpersecution. This necessitated the British 'rescuing them from bondage'in the 'war' of 1835. These Fingos, transformed by their misfortunes,became at once 'the Jews of Kaffirland1, immediately receptive tolabour, Christianity, profit and life as peasant farmers. Fortunatelythis holocaust (nearly two million dead) died down as quickly (andmysteriously) as it began. In the 1830s the whites were able to moveinto empty areas, survey the bleached bones in the veld, wean thecannibals from their habit, and provide a rallying point for thesurvivors. These, nevertheless, remained in the peripheral areas(shaped rather like a horseshoe) - desert, mountains, low-veld - whichprovided the original delineation of the later Bantustans.

To laugh or not to laugh? Students are never quite sure; but thelecturer's cue invariably turns them straight-faced to their notes andexams. It is true that heavy camouflage conceals the blood-line of thisnonsense. Still, how astonishing that it was not questioned before theearly 1980s? The exaggerated teleology and Afrocentricism arenoticeable, as are the pluralistic separation off of black from whitehistory (to remedy this would justify a year's moratorium in our historyexaminations), the whites as incidental and innocent by-standers, andthe over-happy coincidence of the depeopling of Natal, the Orange FreeState and the Transvaal immediately prior to the white occupation. Acolossal, self-induced Auseinandersetzung a split-second before thewhites arrive; as soon as eyewitnesses appear everything is quiet! Afew writers, notably Dora Taylor, Hosiah Jaffe and, more recently,Marianne Cornevin, have shown how aspects of this legitimate apartheid.These texts, however, have not been taken seriously by the Universities.(My own course was unanimously axed when it was seen where my thoughtswere going.) Moreover, nobody has yet questioned the mfecane as totalconcept. Modern Ph.D. students concentrate on 'manageable', i.e.non-controversial issues. A law not unlike that at work at Chernobylhas led, in short, to disaster. A far-reaching inquest is necessary.

The 'perfection' of our mfecane is the product of much accretion, honingand purifying: but the main pillars of the mythology were in place bythe 1830s. By then the literature was already heavily contaminated withfantastical descriptions and fictions composed by settler propagandistssuch as Fynn, Chase and Godlonton. (The style was typical of theperiod, but has only been exposed by Curt in in the west Africancontext.) Behind this ' proto-mfecane' lay the needs to obfuscate theCape's labour procurement strategies in the years after the arrival ofthe 1820 settlers, and to depict Natal as depopulated so as to encouragea northern extension of colonisation. The first produced the myths ofthe Mantatee and Fetcani 'hordes'; the second the fantasy of Shaka andthe equally 'horde-ish' Zulu. Mystifications about the 'gyrations' and

self-inflieted damage of the Mantatees swelled in ratio to the growingamnesia about the slave trade north of the Orange in the early 1820s.The subsequent switch in 1828 to a 'free1 labour strategy produced themassacres of the Ngwane in the same year, condemnation of which wasdrowned in the hypocrisies of Godlonton and his Graham's Town Journal.White destabilizations south of the Tugela and almost certainmanipulation of the Zulu succession in 1828 intensified the anti-Shakacampaign in the early 1830s. The unfortunate timing of a yet huger raidon the Gcaleka for cattle and labour in 1834-5 - a year after Britain'sabolition of slavery - was to necessitate the even more ornate fictionsdepicting Fingos as victims of the Shaka terror. In this manner the twooriginally separate sources of the mythology were fused. Mostincredibly of all, the Mantatee 'horde' was - also in the mid-l830s -elided with the harmless and inconspicuous 'MaNtatisi' , the mother ofthe Tlokwas leader, Sekonyela, an elision additionally connected to alocal Caledon valley propaganda campaign of the British-backedMoshoeshoe against the 'villainous' Sekonyela, his regional rival.

In the half century before 1910 there was a tension between attempts ata pan-South African mythology, and the particularistic local propagandasof the Cape, Natal and, growingly after 1868, Basutoland, a tensionwhich mimicked the pushmipullu-like political tendencies of the era.The Cape after 'Responsible Government' concentrated on embellishing thestory of the Fingos (this reached an apotheosis in Whiteside's Historyof the AmaMbo in 1912) and of drawing lurid portraits of Matiwane.Mbolompo and 'the rescue of the Fingos' became founding battles ofHastings for the lack of anything more appropriate. The 1840s settlersin Natal took up the Zulu mythology. A crescendo of anti-Zuludenunciation occurred to cover the land seizures, hut taxation andchibaro labour systems of the 1840s and 1850s; the attack on Cetshwayo'skingdom in 1879; and the seizure of land within Zululand itself after1897. Cetshwayo was sedulously cloaked in the myth of Shaka; perhapsmore importnatly Shaka was attributed with the armies of Cetshwayo andBarnbata. The myths reverberated back and forth across the decades,producing an analytical timelessness detectable in the traditions ofJames Stuart, whose informants speak with ventriloqual voice. AfterBryant had further worked on this it was difficult not to see the wholetree of South African history as predestined in the acorn that wasDingiswayo. Myth crystalised not merely around the least known men (andwomen) but sprouted most luxuriantly in those geographical regions whichwere the last to be explored by Europeans. Encouraged thus, Ellenberger- in the same era as Stuart and Whiteside - staged his 'lifaqane' in theunknown Caledon of the 1820s: an alleged bloodbath between - precisely -1820 and 1833 organised by Mzilikazi, Matiwane, MaNtatisi and ' thecannibals'. In Ellenberger the hagiography of Moshoeshoe begun by theFrench missionaries in the 1830s reached a peroration. Black chiefseverywhere were thus sorted into 'heroes' or 'villains', as they hadserved, or not, white expansionism.

G.M. Theal was mainly responsible for amalgamating the sectorial mythsto produce a pan-South African history that pre-annexed the BoerRepublics and was the ideological facet of the drive for Union. Thealsculpted even more extraordinary versions on which the text-books arestill inexcusably but logically based. Extra stress was placed onMzilikasi's 'depopulation' of the Transvaal. The central chain-reaction

of ' the mfecane' : Zulu attack Hlubi/Ngwane, who attack Tlokwa (i.e.Mantatees), who expel Kololo, some of whom career on to Dithakong, isnow (1880s) invented. A Zulu-inspired self-genocide of blacks put into'correct perspective1 any damage the whites might have caused. Themagisterial pronouncements of Theal and Cory echo around thedeliberations of the Lagden Commission as they help themselves to 93% ofland in the Union. Maps produced by Theal' s main heir, E. A. Walker,splashed hatenured lines into the central Transkei and Bechuanalanddepicting yet greater swathes of depopulation. In 1928 only did Walkercoin the neologism mfecane (the italicisation an additional disguise) todenote the total process; his translation, 'the crushing', despitepathological repetition in the text-books, has no legitimacy. Theappearance of the Fynn 'Diary' in 1950 and of the largely fictionalShaka Zulu by Ritter established Shaka's world popular reputation,whilst Omer-Cooper's Zulu Aftermath in 1966 improbably established 'themfecane' as a centre-piece of the new 'Africanist' history. The OxfordHistory (1969) injected all this into academic arteries world-wide inthe early 1970s, as a comparison of text-books of that era with those ofthe 1950s will show. The further simplification and caricature of theseversions by television and cinema, and by writers of South Africa'sschool text-books has probably ensured that this mfecane/Shaka issemiologically ineradicable this side of a revolution in our educationalsystem. The failure to see the links between 'the mfecane' of the pasttwenty years and Verwoerd, Vorster and Both's Bantustan strategies -despite its presence on the election platforms of the far right - is notthe least of the items on the agenda of self-criticism for thecontemporary historical profession.

Interlude

My attempts to teach 'the mfecane1 in 1982 (as an optional third yearcourse, paralleling the compulsary South African course!) turned into adetermination to dig up by the roots and expose the very object ofinstruction. I was confronted with an ever lengthening list of errors,omissions of all too easily accessible evidence, and an equallyinexhaustible list of fictional insertions and false ascriptions.Wherever I looked - in 'Natal' , the trans-Mkuzi, or the Caledon - theZulu presence vanished altogether or needed to be rewritten in itscorrect scale and circumstance. Stuart's informants continuallycontradicted his assumptions of a depopulated Natal. No explanationexisted for the first move of the Ndebele; whilst their subsequentpropulsion north was arranged by the Griquas and the Boers. Nothingreliable existed in explanation of any of the Ngwane movements. TheTlokwas had never been reported out of the upper Caledon. A fog ofhypocritical mystification surrounded the 'battle' of Dithakong. Therewas a conspiracy of silence as to the events of 1828 and 1835. Thecircumstances of the 1828 events at both ends of the Transkei had beenshredded, whilst the stories of D'Urban's military of huge migrationsthrough the Transkei seemed a more than convenient alibi forlabour-hungry conquerors. There was a resonant silence as to Portugueseslaving at the Bay: indeed, southern Mozambique had been snipped out ofSouth African history altogether. Repeatedly there had been a'levelling up', so that spear was depicted as more decisive than therifle, minor skirmishes between black groups as dwarfing Austerlitz,Borodino and 'the mfecane' as a whole approximately equal in scale to

the Napoleonic Wars. Settler propaganda had had a century and a half inwhich to manipulate, forget and rearrange the facts. The sacred textsof Ellenberger and Bryant contained no coherent histories at all, merelya bombardment of suggestion, confusingly over-complex detail,patronising pronouncements, chronological inversions, hundredfoldincreases or diminutions in scale, credulous speculation dressed up andaccepted as expertise, the repeated resort to fiction when the threadsof evidence ran out (or were tactfully discarded), and, throughout, themost repellant and racist self-glorification. Nothing of any of thishad been challenged. Stories however surreal, however absurd, had beenhumourlessly repeated by the most respected historians right the way tothe present.

The trouble ahead was illuminated by my failure to answer the firstquestion: what was 'the mfecane' ? It was, for example, impossible todelimit chronologically. Confusion as to the termination sent me in hotpursuit across the borders of South African history 'proper1, aterritory where my questions were not welcomed. Did the Mbolompocampaign, the 'frontier war' of 1834-5 or the Zulu war of 1879 belong tothe course on the mfecane or the one on 'South African history1?Attempts at arbitration evaded the real problem, the untenability ofseparating the material into two courses in the first place. As for themfecane's initiation and causation, the literature was in spectaculardisarray. The fragile attempts of Guy and Hall to provide substance tothe somewhat desperate hypothesis of overpopulation revealed on thecontrary its untenability. The nearly complete vacuum in knowledgeabout Shaka and his Zulu had switched attention back to Zwide in the1790s, and away from the Mfolosi as 'storm centre1 to the Pongola. ButZwide's Ndwandwe were even more of a blank. Hedges attempted to replacethe argument about excess population with one that attributed theinitiation of structural revolution to the supplying of American whalersat Delagoa Bay with cattle in the 1790s. This seemed equallyunsatisfactory. Moreover, this sort of hypothesis formed a differentspecies of explanation which exploded the subject of explanation. Howcould an 'internal revolution' that was isolatedly integral to blacksocieties have been caused by exogamous impacts? Even the militaryrevolution evaporated on close inspection, or, rather, Omer-Cooper'sversion did. The ibutho long predated the 1790s. That it may well havebeen readapted to hunt elephants, cattle (and people?), as Hedges,argued very plausibly, begged a chain of questions. Assumptions aboutthe short stabbing spear and 'horns and chests' tactics were amusing.The fixation with the 'Nguni' was a projection back of twentieth-centurytribe-manufacture. Surely 'Tsonga' and 'Sotho' societies had alsoexperienced dramatic change? Were the 'Nguni1 predators and everyoneelse victims? My conclusion that Mzilikazi's Ndebele and Moshoeshoe1s'Sotho1 kingdoms were sister formations had been anticipated byMacmillan. If one brought in the 'bastard' states, the Taung, the Xhosabands, the raiders of Coenrad de Buys, and the constellations built upby Fynn and Farewell in southern Natal, not only the concept of 'tribe',but even that of 'race' became redundant. Uncoincidentally, this ran upagainst another of the litanies of the 'liberal' world view.

Hedges had advanced to the brink of shattering the mfecane withoutrealising it. My own 'swingeing' onslaught was handicapped by aninitial failure to provide a coherent overview of events that had, at

least some of them, all too clearly occurred. The mfecane was pepperedwith buckshot, but it still lived. The missing bones which permitted afuJ 1 restructuring of the skeleton of the real animal that had stalkedthe sub-continent were supplied by Harries who in 1981 (as spin-off fromhis Ph.D. thesis!) produced incontrovertible evidence of a flourishingslave trade at Delagoa Bay in the first half of the nineteenth century.It was now possible to explain an exodus of peoples in all directionsfrom an epicentre neither on the Mfolosi nor the Pongola, but furthernorth to the west of the Bay. It was immaterial whether this trade hadtaken off after 1815 or whether it had existed (at a lower levelperhaps: we need research) in the eighteenth century. Either way,decisive additional weight was given to Hedge's hypotheses concerningthe transformatory impact of the ivory and cattle trades. Theintensification of regional violence after 1815 was now explanable. thedevious need to resort to the (in fact hypothetical) coming to power ofShaka ' in 1816' fell away. Peoples such as the Ngwane and the Ndebelewere likely to have been expelled from the south-west Bay hinterland bythe slave trade, just as peoples to the north were. But, in thesouthern sector their flight drove them straight into the guns of theGriqua and, later, Boers operating out of the Cape Colony. It was thisdouble pinning, or the simultaneity of antipodal pressures whichdistinguished southern Africa. Both of the slave trades, as well as theupheavals north of the Bay had been removed from history by settlerpropaganda, and Shaka ubiquitously inserted as explanation. The result,'the mfecane', was a contrived illusion of the literature, the negationof 'events on the ground'.

It was this juxtaposition which accounts for the fact that anunparalleled pattern of dislocation occurred in proximity to and, inpart, as consequence of the relatively weak (as compared to centralMozambique or west Africa) slave trade at the Bay. Put the other way,the specific combination of events was driven dominantly by the settlerpresence: without this, black reactions to the local slave trade wouldhave reassembled - perhaps at an equivalently weaker level - thesequences along the Zambezi and the Rufiji. As explained later, theshort initial moves of both Ndebele and Ngwane support this conclusion.The black experience is southern Africa _is, thus, quite different fromelsewhere in Africa, and the nature of the changes within their socialformations has a specific flavour. Nevertheless, such a peculiar andcomplex series of interactions was then thrown comprehensively out offocus by settler writing ('the mfecane'). This misdescription of the1830s and 1840s has, as mentioned, parallels too in other parts ofAfrica. 'Normally1 - if South African had been decolonized after 1945 -a return to history would have been effected long ago. Equally withoutparallel (perhaps in world history), however, is the perpetuation ofthese early sleights of hand and literary habits into the present era ofthe cinema, the television and the paperback; and the readaptation ofmythologies which served one group of purposes before 1850 and thenaround 1900 to new objectives of concealment in the late twentiethcentury.

To Return To History

Extricating ourselves from the quicksands of the Cape-Natal propaganda,we must gain height for a subcontinental view. Observing this with one

eye, we must fly higher yet and inspect the human world as a whole, andover a much longer time-span. We have to concentrate on the twoprocesses at the same time, that in western Europe, 'big1 and 'quick',the other beneath us, 'small1 and 'slow'. And while one eye is fixed oneach, a third must measure the accelerating convergence and impact.Ridding ourselves of assumptions of 'merrie Africa' (whilst reservingthe right to be uneasy about where the now united process is going), andconcentrating on choosing the words with nuances which will offend theleast part of the audience; with a final check on our backgrounds,prejudices and dreams, we can extricate the spare sixth hand and begin.

Out of feudalism in western Europe emerged capitalism, in the firstphase of which the Portuguese and Spanish 'discovered' America, roundedAfrica, and began to harvest the former with the labour of the latter.The fruits returned to Europe to generate new technologies, trades, anobsession with profit, upheavals in 'world view', population explosion,and, sadly, a new era of ever heightening warfare. This phase oftrading/slaving-based, or plunder-based 'mercantile' capitalism producedtitanic struggles between new 'nations', the most strategicallyfortunate and politically innovative of which, England - takingPortugual under a quasi-protection (1654; 1703) and making the Dutchjunior partners (1625-1713), took on the French and by 1800 were halfway to achieving a world empire. Utilising slave labour in theAmericas, and revolutionising land tenure at home, investing the profitfrom the former and juxtaposing it with the internal labour released bythe latter, the British began to industrialise a generation ahead of herrivals. Seeking markets for her depression prone cotton and woolindustries, naval hegemony against the French, and 'temperate1 lands inwhich to settle a perceived (and indeed for a time actual) surpluspopulation - in short lebensraum - the British moved into amongst otherplaces southern Africa (1806), taking the already present Dutch into africtioned tutelage.

On the ground beneath us some very interesting 'interactions' havealready been occurring - and will now intensify - on severalsimultaneous fronts. Around Delagoa Bay, firstly, Portuguese, British,'Austrian1, American and French traders have already by 1800 had aconsiderable impact on local 'Tsonga', 'Sotho' and 'Olentont' societies.Cloth, beads, brass and guns are going in, ivory, cattle, ambergris,gold etc. are going out. There is some uncertainty - c.1800 - whetheryet substantial numbers of slaves are being exported; but in view of thesporadic references to an eighteenth-century trade and of the latenessof our being informed of the very substantial trade after c.1810 we arechecking the evidence very carefully. Following Hedges, we can see thatinterrelationships between states are powerfully affected by thesetrades. Conflict between Mabudu, Tembe and Mattolla is intensified.Ndwandwe and her satelites reorientate themselves to the trade. Furthersouth an Mthethwa-Mabudu coastal alliance is split into two by thewest-south-east axis of Ndwandwe and allies. A glance north of the Bay,however, indicates that trends in state formation and a heightening ofviolence are not confined to the south. Throughout the region olderintra-African trade routes are captured by the new dominant ones endingat the sea. Peoples as far apart as the Pungwe, the Kei and the Molopoare brought within one huge trading network. But all this isunexceptional. It resembles, mutatis mutandis, similar experiences of

8

west African societies in the same era. We have merely lacked a Cur tinto chronicle it.

To the south-west, secondly - and here we come to the 'specificity' ofsouthern Africa - are the Dutch who find no counterpart in west Africa.Between the 1650s and the arrival of the British the Dutch had spreadout and settled the land. They brought new diseases, began the task ofexterminating the San and Khoi, and interacted with the Mozambique coastby importing slaves, and even for a time establishing a trading companyat the Bay (1720s). 'Miscegenation' produced Christianised, gun-armed,horse-riding 'Bastards', who moved into the interior seeking more land,subsumed Khoi groups along the Orange - such as the Kora -, trading,raiding - a reach which extended well beyond the Limpopo by the 1790s.(Overland contacts with the Bay cannot be ruled out.) Remarkable andvery powerful 'commando states' emerged, such as that of KlaasAfrikaaner - which require detailed research. The trading-raiding forivory, cattle and, increasingly slaves - mostly San until the 1810s -had a comparable and simultaneous impact in transorangia1 as thetrading/slaving in the Bay hinterland. Legassick notes thatMoleabangwe's Tlhaping had been 'revolutionised' by about 1810 (acomparison between them and Makhasane's Mabudu or Dingiswayo's Mthethwamight be revealing). These considerations too should wean us from ourfixation with the Mfolosi.

The Dutch, in the third sector, had additionally penetrated east, makingcontact with 'Xhosa1 groups such as the Gqunukhwebe the Ndlambe andRharhabe (='Gaikas') east of the Gamtoos. Trade links were establishedwith the Gcaleka east of the Kei (who also received goods from the Bay).Inevitably there were struggles of slowly mounting intensity over landand cattle. In the late eighteenth and early nineteenth centuries thesefairly balanced interactions ('frontier wars' 1-3) were dramaticallyupset by the supercession of Boer rule by that of the British, whobrought with them the world's most modern weaponry and a 'totalstrategy1 new to Africa. Both the Khoi and the Xhosa worlds fell apart.In 1809-12 the British enserfed the Khoi and began a genocidal campaignto force Bantu groups east of the Fish in order to seize their land.The first relocations over a huge area in South Africa's history wereachieved with a depth of brutality only recently revealed by Maclennan.This 'fourth Frontier War' of 1811-12 began the process of landattenuation and 'Fingoisation'. To speak of •frontier wars' isdeceptive (the euphemism is a favourite one of Cape historians): it wasan unending series of attacks, pressures, subversions, 'treaties1,robberies, cattle seizures, proselytizations, betrayals,misrepresentations, and restructuring. The European God and valuesystem, monagamy, and clothing (to uplift the people of Lancashire andYorkshire) were forcibly introduced, land, labour and cattle seized inreturn. Settler propagandists sedulously minimised the comparativescale of this horror (as compared to the mostly imaginary Zulu'upheavals'), and encapsulated it off both geographically and textuallyfrom relevant events elsewhere in the subcontinent. The near conceptualimpossibility of mentally connecting these events of c. 1810-20 with thecontemporary careers of Dingiswayo and Shaka is a significant triumph ofsettler historiographical suggestion.

Between about 1815 and 1840 there occurred both a growing convergenceand interpenetration between the three sectors of the European advance(as well as the opening up of a fourth between the Mzimkhulu and theTugela), and a mounting intensity of violence, as Europeans seized landin the south, and labour in each of the sectors - destined either forthe Cape farms or the sugar plantations of Brazil, Reunion etc. Thisled to a subcontinental-wide crisis for black and 'Khoisan1 peopleswhich is not remotely comparable to anything that had gone before. Nostatement of this appears in any text book of South African history (ormonograph for that matter). The Bay slave trade was only detected in1981. And the slaving north of the Orange was only clarified in outlinein my own article due later this year (1988) . It is the least ofcoincidences that these years are precisely the years of 'the mfecane'.

In the Cape after about 1810 enserfment of the Khoi, and Griqua raidsfor San never came near to solving a chronic labour shortage. TheBritish ban on the slave trade, the ban on the utilisation of 'kaffir'labour, the intrinsic sparcity of Khoi labour and the impossibility ofpolicing a huge region ensured that the large batches of Britishsettlers arriving from the unhappy Britain of Lord Liverpool had nolegal means of obtaining sufficient labour - an unpropitious fact forthe farming and defending an easterly creeping march. With thecompliance of the Governor the settlers and Boers arranged with Griquaand 'frontier ruffians' to supply them with Tswana and Sotho (i.e.black) labour from north of the Orange. This doubly illegal supplynecessitated the extensive use of euphemisms and lies: the mainly womenand children brought south by force were labelled 'Mantatees' coming 'insearch of refuge1 from an auto-violence organised by their own chiefs!The 'battle' of Dithakong was merely one such raid (in fact on Kwena andHurutshe) for slaves and cattle, unusual only in the extensive evidenceleft by the missionary leaders. The early and mid 1820s thus certainlysaw extensive destabilisations in the region north of the Orange: butthe roots of the violence and of the accompanying hunger lay in thesouth, not in the 'Zulu' east. The flow of violence was from south-westto north-east. The Taung emerged as a predator state - alongside theBergenaars, Koranna and Griqua - in the northern Orange Free State: itwas almost certainly a combination of Taung and Bergenaars which forcedthe Patsa-Kololo north sometime before 1824. Peoples fled in alldirections, enlarging the Tlhaping and Taung, as well as Moshoeshoe'sgrowing state east of the Caledon, and Mzilikazi's on the upper Vaal.Others sought their stolen children in the Cape, ending up as labourersthemselves. The ripples of settler invasions thus had far-reachingeffects on black societies not only east of the Gamtoos, but north ofthe Orange, even of the Vaal. It is artificial to speak of distinct'northern' and 'eastern' frontiers: the Caledon and Kei regions werefused into one interconnected theatre by criss-crossing raiding bandsand fleeing peoples. Xhosa bandits raided north, while groups such asthe Ngwane fled from the Bergenaars south of the Orange, when they wereimmediately characterised by the propaganda of the British and Tembu asa new 'Fetcani Horde'. As the Mantatees before them, the Fetcani becamethe target of British raiding - the Ngwane being massacred in the verymonth of Ordinance 49 (July 1828), which, in response to continuinglabour shortages, further expansion east of the Fish, and the imminenceof the ending of slavery, permitted the utilisation of 'free' black

10

labour for the first time. Mutiliated Ngwane prisoners taken in theraid at Mbolompo became the first 'kaffir' labourers to respond to the'invitation1. Soon they were subsumed under the more versatileeuphemism 'Fingo'. Ingeniously, settler propaganda attributed theregional violence to the Mantatees and Fetcani themselves, a mendaciousdisplacement that converted the captives into Untermenschen, the luckyrecipients of British humanitarian attention. Dithakong and Mbolompobecame heroic vindications of the new order.

Paralleling the labour-raiding in the Colony was a coterminousescalation of the slave trade at Delagoa Bay and Inhambane. This badlyneeds a detailed study, but the following points may be stressed. After1815 an extra demand for sugar in Europe with an accompanying rise inslave prices, an increasing activity of Brazilian and United Statesslavers, and British attempts to keep Portugual from slaving north ofthe equator (treaties of 1815 and 1817) produced a convergence on,amongst other ports, Delagoa Bay. By 1821-22 both Gaza and Mthethwawere present around the Bay, the former at least trading slaves to thePortuguese fort. Tembe, Mabudu and Mattolla were all involved in thetrade. By the later 1820s at least 3,000 slaves - mostly men in thiscase - were being exported annually from both the Bay and Inhambane:this is only the detected number of declared slaves. The slaves wereexported to Reunion, Rio de Janeiro, Havane, Buenos Aires andinnumberable other destinations. These numbers persisted into the 1830sand 1840s, before declining in the 1850s - when exports switched to theBoer farms in the Transvaal. However the figures are read a very highpercentage of males were being seized in the 'Delagoa Bay Hinterland', afact overlooked by every theorist of the mfecane's causation.

It is not yet known where the slaves were taken from. An army of blackmusketeers with Portuguese officers existed at Lourenco Marques for theseizure of slaves; but the details of their activities are missing.Local 'tribal1 warfare in the hinterland was virtually certainlyexcacerbated by the market. Fynn in one of his earlier and moreplausible essays indicates that both the Ndwandwe and Zulu sold theirprisoners into slavery. It raises the question discussed by Curtin forSenegambia to what extent slave supply came from 'normal' wars, or wasfed by raiders stimulated into business by the commercial propsects.Inexplicably Hedges fails to mention slavery at all, although hisarguments for the impact merely of the ivory and cattle trades - whichalso intensified during the 1820s - would have been immeasurablystrengthened. What seems certain is the dominance of 'Tsonga' andNdwandwe-linked groups in the trade. There is absolutely no evidence ofa Zulu hegemony in the Bay area in the early 1820s, or at any time forthat matter. The allegations to this effect were inserted backwards bywriters (or their ghosts) such as W.F.W Owen in the 1830s: theircontemporary accounts of 1823-25 contain no such references. Tshopigroups north of the Bay were among the victims. But there areindications that peoples of the Nkomati, Mbelezi, Usuthu, Ingwavuma,Pongola and perhaps Mfolozi valleys were attacked. The heightening oftension in the subregion can only be explained in this context. Theotherwise inexplicable collapse and dispersal of Ndwandwe itself mustsurely be seen in this light. Whether Zwide himself was a slaversupremo must remain an open question. The mechanics of the triangulastruggle between Mthethwa, Ndwandwe and Mabudu at the turn of the 1820s

11

need studying, but, given the background, it is unlikely that issues ofslaving will be absent. Whether the Ndwandwe-1inked Gaza and Ngoni(Jere) were slaver states and moved into the Bay region attracted by newbusiness opportunities must also for the moment remain hypothesis,although the absence of evidence for a Zulu role in these northerlymovements should be noted. On the contrary, the Zulu movements were tothe south. Whether the subsequent slaving careers of Ngoni groupsbeyond the Save and Zambezi were a logical continuation of their earlierexperiences should also be researched. At the least the absurdassumption of 'mfecane theory' that the Ngoni migrations were impelledby atavistic Zuluism, or even of Zulu attacks, should be discarded. Alast question (for now) is whether the slavery-connected migrations of'Tsonga' peoples away from the Bay, up the Olifants, Letaba and Levubuvalleys that Harries describes for the 1830s - when the slave tradereached a peak - can be antedated to the 1820s or even 1810s. Answersto these and many other questions are blocked by our present lack ofevidence. But at least a search within a promising framework can bebegun.

Turning to the comparable cases of the Ndebele (or Khumalo) and theNgwane: it is obviously significant that contemporary writers connectedthe 'upheavals1 in the Pongola region with the Bay slave trade. In the1820s Macmillan repeated Dr Philip's observation of 1828 that theKhumalo had been evicted by slavers. The Ndebele, on the upper Pongola,and the Ngwane, about thirty miles south-west on the upper Mzinyathi,were not only both well sited for attacks by slavers, but are both knownto have been attacked by the slaver Ndwandwe - and in the moreknowledgeable accounts are depicted as having been expelled by theNdwandwe. The Ngwane were also attacked by the Mthethwa (the precisesequence of these attacks is shadowy, but probably not relevant here)who had an alliance with the slaver state of Mabudu, borrowedMakhasane's musketeers, and are known to have been near Lourenco Marquesin 1821. Conversely, the evidence for Zulu attacks is non existent, atleast for the period before 1830. In both cases there is otherwise ablank in place of an explanation, a blank that I ran into when studyingthe Ndebele in the 1970s. I therefore repeat the argument in my recentpaper, ' Mfecane as Alibi' , that both the Ndebele and the Ngwane wereexpelled wither by primary or secondary slave raiders, and that thetiming of these events is attributable to the rapidly rising exportdemand at the Bay in the years c. 1816-21 and not to the activities ofthe African Napoleon.

It is also crucial to understand, to return to an earlier point, thatthe initial moves of both peoples were short - in the Ngwane case aboutfifty or sixty miles south-west into the upper Wilge, the Ndebele aboutseventy miles north-west to the east side of the headwaters of the Vaalsomewhere near modern Ermelo. (There was, it should be noted, noNdebele migration into the eastern Transvaal.) After further shortNgwane moves from the Wilge into the upper Caledon, the gap between thetwo groups had increased from about thirty to nearly two hundred miles,a distance which was to make the difference between survival anddestruction. In an Africa wide context such flights from slavers were'normal'. Even in the context of the eastern high-veld the incursionsof 'Nguni' from the east was unexceptional per se. What was exceptionalabout these incursions of c. 1818-21 was the scale and the causation. A

12

both the Ndebele and the Ngwane came under attack by theGriqua/Bergenaar raiders operating out of the south-west and searchingfor cattle and 'Mantatees' (i.e. slaves) for the settler farms. BothMatiwane and Mzilikazi were thus caught in the crossfire of the twointer-related plunder systems: both were caught and kept moving by asecond 'system'. Mzilikazi's people, about 400 miles/600ks from theGriqua bases!, were in a more favourable position, but were nonethelessseverely harrassed in the mid 1820s by Griqua attentions. The exactmoment of the Ndebele migration out of the south-eastern Transvaal intothe western Transvaal is still open to question (some time between 1827and 1833), but the Griqua causation is undisputed. The Khumalo stateexpanded by the absorption of 'Sotho' and 'Tswana' refugees displaced inthe southern destabilisation - people fleeing the Taung and Bergenaars -on the one hand, and 'Nguni' (e.g Ndwandwe) and possibly 'Tsonga1 groupsdisplaced like themselves in the eastern displacements on the other. Itwas in these circumstances of a sequence of harrassments that thepeculiar evolution of Ndebele amabutho began. The Ngwane, conversely,were by 1824 much more exposed to Griqua attacks where they had arrivedwest of the upper Caledon - far more endangered than Moshoeshoe's morepropitiously sited Mokhoteli in the mountains east of the river.Emulating the Ndebele in some respects, the Ngwane were shattered byGriqua attacks in c. 1825-26. The following migration through thenorth-eastern Cape into the southern Transkei - the only direction ofescape open - ended in a third disaster when they were massacred by theBritish army in July and August 1828 in the immediate aftermath of thesanctioning of 'free1 labour in Ordinance 49. The Ngwane were flunghelplessly between the three penetrative fronts of white activity. Boththe teleology and Afrocentricism of mfecane theory collapse.

With some of the ground cleared, it is safe, and chronologicallyapposite to turn to the early Zulu. The general failure to probe thefictions of Owen, King, Farewell, Fynn, Chase and Godlonton, and themost uncritical use of Stuart and Bryant, has ensured that - with thenotable exception of Hamilton and Wright's recent studies - the analysisof the pre-1840s Zulu kingdom is a disaster area. The first thing tonote is that what linked the three theatres of Delagoa Bay, 'Natal' andthe trans-Kei was not Shaka's impis but settler propaganda. Thwarted atthe Bay by the failure of the British Government to accept Owen's 'TembeTreaty' of 1823 (as well as by malaria), settlers landed at Port Natalin 1824 and at once proclaimed: a recent Zulu devastation of the Mpondo;a Zulu depopulation of Natal; and a land treaty granted by Shaka to thearea so generously depopulated. During 1825-27, as an aphrodisiac forthe scarcely interested Cape administration and merchant houses, avilification campaign was unleashed against Shaka and the Zulu. Everyexhortation, slander, deception and exaggeration was deployed in aninitially unsuccessful attempt to tempt Britain north. Zulu tyrany overthe Bay hinterland was also invented in case British chances in thatarea took a turn for the better.

Not one of these charges stands up to examination. There was nodepopulation of Natal, no Zulu hegemony at the Bay, no 1824 land treaty,and, virtually certainly, no Zulu attack on the Mpondoin 1824. Neitherwas Shaka the monster depicted in the pages of Fynn and Isaacs. Ifanything, the signs are that as a 'reactive' state (or 'defensive' stateto use the terminology of Wright and Hamilton) the Zulu were in the

13

1820s as yet comparatively weak. The emergence of Zulu power after thedemise of Mthethwa (c. 1815-23) must be seen both in the context of theslave trade and the opening of a second European front at Port Natal.The first Zulu amalagamations occurred along the White Umfolosi.Peoples along the Mkuzi and the upper Mzinyathi - such as Hlubi andNgwane, even Ndwandwe groups - fled into the growing state. The1Ndwandwe-Zulu' fighting so precisely (and erroneously) dated to 1818-19by Bryant (the repeating of Bryant's guessed datings is a sympton of theliterature) is 'a myth' - probably indicating not a sequence of specificbattles, but a reference to the generalised upheavals between thePongola and Mfolozi. (And what were the Ndwandwe doing when theyattacked south of the Mfolosi in '1818-19' and drove 'the zulu1 to theTugela?) In this regional struggle the Zulu-linked units were worstedand at the turn of the 1820s the centre of the growing Zulu stateshifted south-east into Qwabe territory, leaving military units to thenorth of the black Mfolosi in a defensive ring. From 1824 the Zulu,like the Mabudu in the far north, received the aid of white gunmen, whoin 1826-7 enabled 'the Ndwande' to be attacked and Khumalo groups alongthe Mfolosi forced to khonza. These 'victories' in fact merely indicatethe weakness and unexceptionality of the Shaka state until that date.They were followed by a further shift of the Zulu centre southwards, thecapital itself moving south of the Tugela in 1825. If anything the axisof the state was at this stage along the Tugela, as Cele and Mbo groupsin the region voluntarily fused with Shaka. Whether these southerlymovements are to be viewed as a reaction away from the slaving areas, oras positive alignments towards the Europeans at Port Natal is an openquestion. To attempt to explain these sequences in terms of a searchfor a better combination of grazing and cultivation fails to provide anadequate balance between cause and the issue (the growth of the Zulustate) to be explained.

It should be stressed that whilst the evolution of this state, asHamilton makes clear, is complex, the scale of the changes should not beexaggerated. The formation associated with Shaka is comparable withother emergent states of roughly the same era such as Mzilikazi's andMoshoeshoe's, both equally complex, equally 'revolutionary', equallyassimilative. Additionally, Shaka's state was in the context of Zuluhistory merely the first and most primitive stage in an evolutionarysequence which was to culminate in Mpande' s kingdom of the 1860s. Inview of the regional circumstances in which the Zulu kingdom developedbetween about 1815 and 1830 it is surely misleading to term it a' precolonial' state at any point in its career. Its very nature wasinfluenced by white pressures and intervention. The whites themselvescreated mini-states in the area to the south of the Port which changedfrom being allies of the Zulu to profound threats. In 1828 it was analliance of these white-led predatory groups which attacked the Mpondo.Ordinance 49, the raids on the Ngwane, and Fynn and Farewell's attemptsto entice settlement north led to a reckoning in which the whites joinedwith internal opponents to kill Shaka in September 1828 and install thehopefully more malleable Dingane. It is significant that some of themost powerful of Shaka's supporters refused to accept Dingane, and thatan extensive civil war occurred between 1828 and c. 1832 in which thewhites backed Dingane. Dingane's victory, the transference of the stateback north of the Tugela, the slaving activities of Farewell, Fynn andIsaacs, the falling out of the whites and Dingane, the adaptations in

14

state structure during Dingane's career, the extent and nature of Zuluinvolvement at the Bay and as slavers in the 1830s are all issues whichhave so far remained unstudied. By another irony, explainable by thelater historiographical manipulations, the post-1830s development of thekingdom under Mpande and Cetshwayo produced something like thepropaganda picture of the 1820s - another conjurer's illusion which'confirmed' the early literature and closed it to question.

Possibly even less researched than any of the foregoing are interactionsin the region between the Kei and the Mzimkhulu during the 1820s. Therepeated insinuation of 'mfecane theory' that this area was penetratedby huge numbers of refugees fleeing Shaka has never been seriouslyinvestigated and is not supported by the evidence. On the contrary, the'transkei', like the Caledon, was a regional interface between thefronts of white penetration - within reach of refugees from thesouth-west Bay hinterland, the Caledon and Orange, and the easternfrontier. Local predators such as Faku's Mpondo, Ncaphayi's Bhaca andNgugencuka's Tembu were joined by Fynn's own raiders, by white and'coloured' hunters, traders and bandits, by the cross Drakensberg raidsof the Mokhoteli and Tlokwa, and by the incursions of the Ngwane. Inthe north-west, Hlubi incursions are as likely to have had their originin slaving as in Zulu action (and note the Hlubi who fled into Shaka'skingdom). Most lethal of all were the compressions of the British inthe south. In 1826 British missionaries crossed the Kei into Gcalekaterritory. A crossfire of propaganda between the Mzimkhulu andButterworth pleaded for direct expansion north of the Kei through Mpondointo Natal; to further this, alliances were made with both Mpondo andTembu first against the Ngwane, subsequently against the Gcalekathemselves. Both groups were caught between two fires of whitedestabilisation; but we have heard that before. All this awaits astudy; nevertheless the primary of southern pressures is clear. Whatpassed south was British disinformation, not Zulu armies and refugeesdisplaced by them.

The first people labelled 'Fingo' were the odd individuals whogravitated to the mission stations set up east of the Fish river in theearly 1820s. There are some indications that the word had southernorigins west of the Fish, and that its versatile meaning embraced earlyinitiates, tribal misfits, opportunists and 'collaborators', displaced' kaf f irs' in the region south of the Kei, as well as migrants, forexample from north of the Orange. In missionary useage around 1828'Fetcani' and 'Fingo' are interchangeable in many passages. One doesnot have to look far for the agents of this uprooting and labelling.The Ngwane captured in the raids of 1828 were also designated Fingos, anearly use of the term for 'forced labourer'. In the early 1830s theword was being used for people as widely divergent as forced labourersand a contingent of armed Xhosa collaborators who were to be used in thenext 'frontier war'. The accretion of meaning; the very fuzziness ofthe word was to have advantages after 1835.

By 1833 an even acuter shortage of labour on the eastern frontiercoincided with Britain's abolition of slavery. This did not deter thesettlers from raiding the Rharhabe for slaves in the period before thewar. But even more elaborate covers as compared to those of 1828 werenecessary. The 'war' that followed, 'D'Urban's war' of 1834-35, was in

15

essentials a massive land, cattle and labour raid on the Rharhabe andGcaleka in particular, which dwarfed the Mbolompo campaign in scale.The British army seized well over 50,000 'Xhosa' cattle, and, a factunnoticed even by specialists on the Gcaleka, 17,000 (sic) Gcalekaprisoners - 85% of whom were women and children - in a series ofspectacular raids in the first few months of 1835. A combination of the'war' and the penetration of the region between the Keiskamma and theKei by British administration at last solved the labour supply crisis atthe crude level. The further land seizures over the next twenty years,masters and servants ordinances, hut taxation, and missionaryencouragement soon fine-tuned a modern labour procurement strategy whichduring the 1890s spread throughout Africa. Given the context of 1835,however, the true provenance of the Fingos, as the prisoners inevitablywere called, was concealed. The raids around Butterworth and along theupper Kei of March-June 1835 were erased from the texts. An elaborate,though threadbare, story of vast groups fleeing from the Zulu only to bepersecuted by the Gcaleka - an updating of the 1828 Zulu raid alibi -was unimaginatively evolved by D'Urban's military entourage, which,suspiciously, included H.F. Fynn. It is not yet known where the Ziziand Bhele 'chiefs' were obtained, but the possibility that they weremembers of Fynn's Mzimkhulu hierarchy, temporarily unemployed, is thehypothesis under investigation. As for the Hlubi 'Fingos', they werepeoples who migrated into the north-eastern Cape from the eastern OrangeFree State in the early 1850s and after, who were inserted back into the1 flight from the Zulu in the 1820s' in subsequent historiographicaladjustments. The fingo fiction satisfied 'London' with one and a halfeyes averted. Throughout the 1840s and 1850s the false history,originally attributable to the military, was embroidered byintellectuals and missionaries. The Butterworth missionary, Ayliff, forexample, whose contemporary diary of 1831-34 makes no mention of theseFingos, plagiarised the settler-military myth in a series of essays inthe early 1850s, which were paraded by later writers as if they had beenwritten in the early 1830s. This in turn, along with much laterfantasies about ' the Mbo' penned by Scully, was incorporated intoWhiteside's History of the AmaMbo (1912). The uncritical repetition ofthis material by historians of the eastern Cape is worth an inquest initself. Of course, the spectacular increase in the population of Fingosbetween the 1830s and Union has nothing to do with sexual reproduction,but quite a lot to do with displacements in subsequent 'frontier wars',hut taxation, accelerating peasantisation, and adjustments to image inthe schools. 'Xhosa' collaborators who crossed north of the Kei intoFingoland in the 1860s took with them the baggage of a false history aspart of the price for the land.

It is time, too, to end the debate as to the relative centrality and'creativity' of the 'Great Trek1 and 'the mfecane'. Whilst the formeroccurred, the latter is an historiographical illusion. They are facetsof the identical process. The Boer invasions of the high veld after1835 were now to trigger events which were continuation andintensification (as well as modernisation) of previous pressures, as thetrek itself was merely a stage in European expansion. The collision atVegkop (1836) and the slaughter of Retief's band (1338) coloured andheightened the myths about the Ndebele and the Zulu which the Boers,departing from Godlonton's Grahamstown, took with them in the firstplace. Black states grew into their own myths. Like the Ngwane in

16

1828, the Ndebele were now hit by a third and infinitely stronger force:an alliance between the Boers, Griqua and both 'Sotho1 and 'Tswana' fromthe far from depopulated Orange Free State. After bitter fightingduring 1838 the Ndebele fled north of the Limpopo. Dingane waseliminated two years later, Mosega and Blood River indices of the truebalance of power in the subcontinent. Mpande's kingdom of the 1840s and1850s was an opportunistic response to the particular outcome of theAnglo-Boer struggle south of the Tugela. Moshoeshoe's preciselycontemporary kingdom east of the Caledon had analogous foundations.Both 'Sotho' and 'Zulu' kingdoms were defined in terms of boundariesduring a lengthy struggle to prevent themselves caving in completely.Both kingdoms reached their peaks in the 1850s, the former collapsingunder Boer attacks in the 1860s, the latter succumbing to direct Britishinvasion in 1879. To repeat an earlier point: at no time can either ofthese formations be termed 'pre-colonial'. It was the pressures ofcolonial expansion and slaving which produced them in the first place.Modern concepts of 'pre-colonial societies' and 'Africanist history1

have closely accompanied the concept of the mfecane, and have servedeven when unintentionally, the same type of ends.

Similar patterns of state evolution occurred during the era of the Boerinvasions of the northern and eastern Transvaal. Albasini's state onthe Letaba, for example, drew peoples fleeing the slave trade at theBay. In certain respects it paralleled the mini-states set up by Fynnand Farewell on the Mzimkhulu a decade earlier. The Maroteng-Pedi andDlamini-Swazi 'reactive' states effloresced (after relatively fortuitousbeginnings) in the 1840 s, reached a peak in the early 1860s, andprogressively collapsed after about 1875. Maroteng state building goesback at least to the 1770s. Delius notes a general interconnectionbetween an increased regional violence at that time and the trade at thecoast. Not having information about the slaving, however, he quiterightly remained puzzled by the lack of sufficient cause for trendswhich extended to the Steelpoort. The Dlamini fled from the south Bayarea across the Lebombo into the middle Pongola valley, perhaps sometimebetween 1750 and 1790. The bids to produce 'mfecane' genealogies forboth states is somewhat artificial, Delius resorting to a possibleNdwandwe invasion of the Steelpoort valley in c. 1822, Bonner, for theDlamini, to a hypothesis that Swazi militarism derived from Zwide viaSobhuza's Ndwandwe wife, Fulata. Without the Boer invasions of the1840s and the slaving at the Bay, however, there would have been nothingto write about. As with his competitor Albasini, Sekwati (Pedi)augmented his following with peoples who had fled both from the Boersand from the eastern low veld, as the Gaza stepped up slave raidingafter about 1842. As with Moshoeshoe in the Caledon, Sekwati had toshoulder aside local rivals: who was to create the nucleus of the local'reactive' state remained in doubt until fairly late. More spectacularstill was the expansion of Mswati's 'Swazi' state in the 1850s.Declining slave prices at the Bay induced Manukosi of the Gaza to switchsupply to the eastern Transvaal Boers - with payment mainly in guns andhorses - whose demand for labour reassembled that of the Albany settlersin the 1820s. Mswati's successful challenge of the Gaza monopoly wasaccompanied by far-reaching military restructuring in the 1850s. Thestate began to swell into its later boundaries with the absorption ofthe hitherto independent amakhandzambile ('those found ahead'), again inthe 1850s. Nothing of this was predestined in Sobhuza's reign.

17

'The mfecane' thus breaks down in every one of its sectors, whethergeographical or chronological. The whole is rotten; and so is each ofthe parts. The attempts to 'reform1 the mfecane by Maylam and Davenportby conceding the point about depopulation, or contending that this orthat was exaggerated (perhaps a little rephrasing...) are doomed tofrustration. (There is no half-way house). The thing is unreformable,unadjustable, unrepairable. Keep it (if you like) or abandon italtogether. In the latter case an enormous hole opens the filling up ofwhich should keep us busy for many years (if we have those years).

In my remaining paragraph I must ul tra-briefly turn to the problemsinherent in defining and classifying the new type of states whichemerged during this {longer: c. 1740s - 1850s) era. They were notspecific to southern Africa at all, but resembled (and differed from)contemporary formations in both east and west Africa. The changes didnot revolve around the emergence of a 'regimental' or 'raiding'formation: that is a legacy of the Zulu-centric approach, an error whichfollows from swallowing 'the mfecane'. Yet the Zulu was a variantexample of the general process. It was not the emergence of the ibutho,but the restructuring and readaptation of the ibutho in the context ofEuropean pressures which was significant. The 'revolution' was not even•Nguni'. It was not even 'Bantu1. Parallel and equally far-reachingchanges took place within 'Tsonga' , 'Sotho' , 'Tsv/ana' , and, for thatmatter, 'San1 and 'Khoi' formations. Given the fusion of differentpeoples in many of the examples, these terms become a hinderance toanalysis. There was such a wide spectrum of equally relevant 'emergent'states, with 'war lord' leaders (to use Martin Hall's term), that it isdifficult to know where to place the boundaries. Note the rainbowsequence: Boer commando - Griqua commando - Koranna - Taung - Kololo -Ngwaketse for example. Or, British army - Fingo units - Ndlambe - Tembu

- Mpondo - Gcaleka. Or, Fynn and Farewell's organisation - Thuli - Cele- Mbo - Zulu. Alternatively, Albasini's state - Sekwati's - Pedi -'Swazi' - Gaza - Portuguese slavers - to go round in a circle. Innearly every instance there were increased power of chiefs, newhierarchies, new patronage systems, restructuring of militaryorganisations, the obtaining of firearms, larger size, amalgamations,greater levels of violence etc. To attempt to establish a general 'law'for all this - a sort of mathematical formula - is surely to chase amirage. None of these adapted states was 'pre-colonial' at any moment:the idea of 'a precolonial mode of production' (southern Africanvariant) is also a fantasy. Each, if I may make my one reference toPerry Anderson, is definable in terms of its 'superstructure'. Thisgives us greater analytical flexibility, and enough oxygen to be able toreturn to history and submerge for long enough to be able to retrievesufficient empirical material free from the worry about having topronounce a general law. A prize for the one who discovers anunsullied, 'pure' African state, with a real, uncontaminated 'internalrevolution'. The mere existence of evidence precludesnon-contamination. And with no evidence the secret of the 'innertrajectory of change1 will go to the grave in silence. Two things aresurely certain: in origin all were 'defensive', or 'reactive' even,indeed, especially when successfully expansionist; and all of them,brought into being by capitalism, embracing what they could ofcapitalist techniques in order to avoid catastrophe, were, nevertheless,one by one, ununitedly, broken down and exploded by the system which had

18

jolted them into life.

In Briefest Conclusion

The Afrocentricism, Zulucentricism, and both the chronological andspatial teleological chain reactions of 'mfecane theory' collapse onclose inspection. External pressures - particularly the heighteneddemand fo sugar in Europe and British market and lebensraum requirements- involved change over a much lengthier period than the (in thiscontext) irrelevant boundaries of Shaka's reign. The Zulu were only oneof a spectrum of 'reactive' states, some of which experiencedaccelerated change in the eighteenth century, others only developingafter Shaka was dead. There was never either a Zulu or 'Nguni' centreto 'the mfecane'; rather, there were several peripheral but convergingfoci of white penetration: the Cape, 'transorangia1, Natal and the Bay.Black societies were like fish caught in a net: flight from one pressureinvariably took them straight into another perhaps worse (the Ngwanenightmare being only the most notable). This accounts for the paradoxthat the relatively unexceptional slave trade of southern Mozambiquecame particularly in the period after 1800 to be interinvolved inconcatenations and escalations of violence which are unusual in Africa.Whereas the 'normal' slave trade was critically amplified by settlerexpansion and labour procurement systems, memory of the slave trades waserased by settler propaganda. Removing themselves from the scene oftheir own impact settlers achieved an historiographical sleight of handwhich was perfected over one hundred and fifty years. This vanishingwas accompanied by the unscrupulous framing of Shaka and other'innocent' black figures such as, wierdly, MaNtatisi. The initial alibiof concealing labour raids of the period c. 1820-1870 was overlaid byanother in the early twentieth century misaccounting for the landdivision of 1913. A post 1940s variant depicts the Bantustans asresulting from a self-sequestration by blacks into the areas they occupytoday during the 1820s and early 1830s. The longevity of such compoundlies was assured by the failure of decolonisation in the aftermath ofthe Second World War, and by the deceptive normality and quality ofSouth African historiography during the 1960s and 1970s. It was not,therefore, 'the mfecane' that was unique to the world; it was, rather,the unsifted preservation of the flamboyant fictions, fantasies and liesof the era of Fynn, their ornamentation by Theal and his colleaguesbefore the First World War, and their delivery intact to the media,educational and propaganda machinery of one of the most efficient andfuturistic of totalitarian states of the present. The resultant evenfurther embroidered mythologies and lies are fed to our schoolchildrenand university students daily.

POLITICAL MYTHOLOGY AND THE MAKING OF NATAL'S MFECANE

John Wright

Department of Historical StudiesUniversity of NatalPietermaritzburg

Introduction

Over the last twenty years or so the concept of the mfecanehas come to be deeply rooted as a notion round which much ofthe history of southern Africa in the first half of the 19thcentury is written. As generally used, the term refers to aseries of wars and migrations which are supposed to have beensparked off by the emergence of the Zulu kingdom in the late1810s, and then to have swirled across most of the easternhalf of the sub-continent. In the view of many historians,these upheavals were the direct cause of the profound changesin the political map of southern Africa which took place Inthe 1820s and 1830s, changes which in turn were of thegreatest significance in shaping the nature of black-white in-teraction in southern Africa for the rest of the century. (1)

In a series of so far unpublished papers written since 1983,Julian Cobbing has formulated a radical and sweeping critiqueof the notion that the mfecane actually happened. (2) Whilenot denying that the history of African societies in the ear-lier 19th century was marked by numerous violent conflicts, herejects the particular significance which white writers sinceat least the mid-19th century have attached to them. Heempasizes that they were a continuation of conflicts which hadbegun long before the 1810s, conflicts whose primary cause wasnot the expansion of the Zulu kingdom but the onslaught whichDutch and British settlers and imperialists at the Cape and,to a lesser extent, Portuguese slavers at Delagoa bay weremaking on neighbouring African societies in their unrelentingattempts to seize control of land and labour-power. Theupheavals of the times had not one but several epicentres. The

1. For example J.D. Omer-Cooper, The Zulu Aftermath: aNineteenth-Centurv Revolution in Bantu Africa. London, 1966,chs. 1, 12; P. Maylam, A History of the African People ofSouth Africa, Cape Town, 1986, ch. 4; R. Edgecombe, 'TheMfecane or Difaqane', in T. Cameron & S.B. Spies, eds., An Il-lustrated History of South Africa. Johannesburg, 1986, ch. 9.

2. J. Cobbing, 'The case against the mfecane1, unpublishedseminar paper, University of Cape Town, 1983; 'The caseagainst the mfecane', unpublished seminar paper. University ofthe Witwatersrand, 1984 (a slightly modified version of theUCT paper); 'The myth of the mfecane1, unpublished seminarpaper, University of Durban-Westville, 1987; 'The mfecane asalibi: thoughts on Dithakong and Mbolompo', unpublished ms.,1987.

role attributed in the literature to the Zulu is not based onhistorical evidence: rather, it is a product of the searchmade by imperialist and settler ideologues for a plausiblealibi for the colonial-based interests whose aggressions wereultimately responsible for the violence and social disruptionsof the period. For their own various ideological reasons,subsequent generations of historians, including that of thepresent, have either been concerned to maintain the alibi, orat the very least, have done nothing to demonstrate its fal-sity. From this perspective the notion of the mfecane is noth-ing but an interest-serving myth. Historians, Cobbing argues,urgently need to abandon not just the term itself, but thewhole set of interlinked assumptions, distortions and false-hoods which it embodies, and to address themselves to thebusiness of developing an entirely new analysis of southernAErican history in the late 18th and early 19th centuries.

This paper, which focusses specifically on the Natal region,lends support to Cobbing's notion of the mfecane.as myth. (3)It presents, in bald outline, the results of a survey of ac-counts published since the late 1820s of the upheavals whichare supposed to have taken place in Natal south of the Thukelariver in the late 1810s and early 1820s. In these literarydescriptions, as they have emerged over a period of 160 years,three elements have been found to recur. First, that Natalwas 'devastated1 (to use one of a group of synonyms which ap-pear in the literature with monotonous regularity) and itspopulation largely exterminated or driven out. Second, thatthe devastation was the work of the Zulu. Third, that it wascarried out as an act of deliberate policy. Over time, thehistorical context of what is here called the devastationstereotype has varied, but the three elements which make upits content have remained firmly fixed. In its entire history,only one writer has made more than a token attempt to root itin historical evidence. (4) From a historiographical perspec-tive, the mfecane as it is supposed to have happened in Natalturns out to be the latest reworking of a set of ideas whosehistory dates back directly to the late 1820s and early 1830s.Generation after generation of writers has uncriticallyreproduced these ideas to form one of the most enduring mythsin southern African history-writing. Before the history ofNatal in the early 19th century can be rewritten, the povertyof mfecane-theory's intellectual lineage has to be laid bare.This paper aims, by outlining an explanation of how thestereotype came into existence, and why it has survived for solong, to contribute to that process.

3. The argument that follows is a heavily compressed versionof a much longer study written as a chapter of a not yet com-pleted Ph.D. dissertation on the history and historiography ofthe Thukela-Mzimkhulu region in the late 18th and early 19thcenturies. My thanks go to Julian Cobbing and Philip Bonnerfor their critical comments on the fuller version.

4. This was A.T. Bryant: see pp. 10-11 below.

The stereotype created

The history of the devastation stereotype begins soon afterthe arrival in 1824 of the first party of British traders toestablish themselves at Port Natal. Within a short time oftheir arrival they were beginning to report that the regionround Port Natal was largely empty of population, and to de-velop and publicize an explanation for this phenomenon. Thisexplanation, to the effect that the previous inhabitants hadeither been killed or driven out by the Zulu under Shaka a fewyears before, was beginning to appear in Cape Town newspapersin 1825 and 1826, and in printed books by 1827. (5) At thesame time specific literary images of the Zulu and of Shakawere beginning to take shape, with the Zulu being described bywriters like King and Thompson as the warlike and bloodthirstyagents of Natal's devastation, and Shaka as the ferocious andsavage leader who directed them. (6) Over the next decadethese ideas were consolidated by a number of other writers -Owen, Pringle, Kay, Boteler, Steedman and, in particular,Isaacs - into literary forms which in.their essence have sur-vived to the present day. (7)

Few of these writers had actually set foot at Port Natal, andnone of them had been eyewitnesses of the processes ofdestruction which they adumbrated. The evidence on which theybased their descriptions was derived directly or indirectlyfrom Africans living in Natal, but nowhere in the works ofthese writers is there any mention of the identity of theseinformants, or of the circumstances in which their testimonywas obtained and recorded. Most of the information which wasdistilled into the devastation stereotype was probably col-lected from members of the remnant groups which the tradersfound living about Port Natal and in the neighbouring coastalregions. It is germane to make the point here that Port Natal

5. B. Leverton, ed., Records of Natal, vol. 1, Pretoria, 1984,p. 51; G. Thompson, Travels and Adventures in Southern Africa,ed. V.S. Forbes, Cape Town, 1967 (1st ed. London, 1827), vol.1, p. 172, voi: 2, p. 249.

6. Thompson, Travels and Adventures, vol.1, pp. 172, 174-5,vol. 2, pp. 248, 249.

7. H.W. Robinson, ed., Narrative of Voyages to Explore theShores of Africa. Arabia and Madagascar Per formed... under theDirection of Captain W.F.W. Owen, R.N., vol. 1, Farnborough,1968 (reprint of 1st ed., London, 1833), p. 71; T. Pringle,African Sketches, London, 1834, p. 362n; S. Kay, Travels andResearches in Caffraria, New York, 1834, pp. 281, 341, 343,344; T. Boteler, Narrative of a Voyage of Discovery to Africaand Arabia, vol. 2, London, 1835, p. 303; A. Steedman, Wander-ings and Adventures in the Interior of South Africa, vol. 2,London, 1835, pp. 200-201; N. Isaacs, Travels and Adventuresin Eastern Africa, ed. L. Herman & P.R. Kirby, Cape Town, 19 7 0(1st ed. London, 1836), esp. ch. 18.

was situated precisely in the one region of Natal which had infact been overrun by Zulu forces, (8) and it is likely thatthe generalized depictions of the destruction of Natal whichwere noised abroad by the traders were a reflection of theparticular experiences of informants from this region.

For their own part, the traders had every interest inpropagating the idea that Natal was, by the time of their ar-rival, largely empty of inhabitants, and that Shaka's Zulu hadbeen the agents of their dispersal or destruction. From thevery first, the leading traders at Port Natal were concernednot only to open up commerce with the Zulu but also to try toestablish rights to large tracts of land round the port. (9)In the face of the express reluctance of the Cape governmentto sanction the acquisition of further territorial posses-sions, (10) they attempted to minimize objections to theirproceedings by asserting the claim that the land in questionwas virtually uninhabited.

If the Port Natal traders had a direct material interest inpropagating the 'myth of the empty land1, so too did the Capemerchants who in large part financed their early trading ven-tures. (11) In the late 1820s and early 1830s the rising com-mercial class and its associates in both the eastern and thewestern Cape were beginning to exert pressure on the Britishauthorities in Cape Town and London to annex Natal and estab-lish it as a colony of British settlement. (12) This class'sspokesmen often used the notion of a depopulated Natal to un-derpin their arguments for the desirability of the territory'sannexation. (13)

8. This emerges from a piecing together of the evidence con-tained in A.T. Bryant's works (see pp. 10-11 below) and in thefour volumes so far published of The James Stuart Archive ofRecorded Oral Evidence Relating to the History of the Zulu andNeighbour ing Peoples, eds. C. de B. Webb & J.B. Wright,Pietermaritzburg, 197 6-86

&

Records

9. H.F.Fynn, The Diary of Henry Francis Fynn, ed. J. StuartD. McK. Malcolm, Pietermaritzburg, 1950, pp. 86-8; Isaacs,Travels and Adventures,, pp. 142, 180-1; Lever ton, ed. , Recoof Natal, vol. 1, pp. 37-40, 247-8.

10. Leverton, ed., Records of Natal, vol. 1, p. 36.

11. On the financing of these expeditions see Fynn, Diary, ch3; B. Roberts, The Zulu Kings, London, 1974, pp. 8-19, 75-6,78-81.

12. S. Bannister, Humane Policy; or Justice to the Aboriginesof New Settlements. appendices 1, 6 & 7; J. Chase, ed. The Natal Papers, Cape Town, 1968 (1st ed. Grahamstown, 1843), pp.23-31; P.R. Kirby, ed., Andrew Smith and Natal, Cape Town,1955, pp. 5-7, 145-6; Roberts, The Zulu Kings, pp. 222-5.

13. Kirby, ed., Andrew Smith and Natal, pp. 148, 155, 170; L.Herman & P.R. Kirby, 'Nathaniel Isaacs: a biographicalsketch', in their edition (cited at note 7 above} of Isaacs's

Propagation of the myth thus served a clear material purpose.So too did the fostering of the image of Shaka as the crueland despotic leader of a warlike Zulu nation. Though in writ-ing of Shaka and the Zulu in lurid detail, writers like Kingand Isaacs clearly had an eye on their reading public, theyand others also wrote to publicize and propagate the pro-annexationist cause. By depicting the Zulu and their king as apotential threat to the security of the Cape's eastern fron-tier region, or alternatively as the potential allies of rivalpowers, they hoped to influence the British authorities intoannexing Natal and thereby paving the way for the extension ofBritish trade and settlement. (14) Though some scepticism wasvoiced in the Cape Town press about the reality of the imageof Shaka put about by the Natal traders, (15) in the early1830s the notion of the 'numerous and warlike1 Zulu as a'threat' was becoming widespread in the discourse of Cape mer-chants and others with an interest in Natal. (16)

By the mid-1830s, then, the devastation stereotype was wellestablished in the literature on southern Africa. It is im-portant for the argument being advanced here to note that ofthe writers so far cited as having been responsible for fixingthe stereotype in print, virtually all were either members of,or had close associations with, the business communities ofCape Town and the eastern Cape. (17) All of them directly orindirectly propounded the virtues of the regions beyond theborders of the Cape colony as fields for the expansion of Capeand British commerce.

From the later 1830s the stereotype began to be taken up anddisseminated by members of another influential body ofopinion-moulders and image-builders, the missionaries who wereactive in southern Africa. Though there were wide differencesof opinion between them as to the benefits or otherwise of theextension of European settlement, they were united in wantingto promote 'civilized' European government. Many aligned them-selves with the .merchants' call for the 'opening up1 of theinterior, and at the same time used much the same kind of his-

Travels and Adventures, p. xi; R. Godlonton, IntroductoryRemarks to a Narrative of the Irruption of the Kafir Hordes,parts I & II, Cape Town, 1965 (1st publ. Grahamstown, 1836),pp. 165-6.

14. Thompson, Travels and Adventures. vol. 1, p. 174, vol. 2,p. 2 49; Isaacs, Travels and Adventures. p. 3 39.

15. Roberts, The Zulu Kings, pp. 154, 177.

16. Chase, ed., Natal Papers, p. 27; Godlonton, IntroductoprvRemarks, pp. 162-8, 172; Kirby, ed., Andrew Smith in Natal,pp. 149-51, 153-4, 166-8, 171-2.

17. The evidence is presented in detail on p. 14 of the widerstudy mentioned in note 3 above.

torical rationale to justify it. (18)

The stereotype appropriated

In the 1840s the hopes of merchants and missionaries for the•opening up1 of Natal at last began to be realized. In 1843Britain annexed the region between the Thukela and theMzimkhulu, and in 1849-51, several thousand British settlerswere established in the new colony to help speed up the pro-cess of 'civilizing' it. But with the achievement of some ofthe main goals of its progenitors, the devastation stereotypedid not wither away. On the contrary, over the next fewdecades it was vigorously taken up, embellished and propagatedby numerous writers in the service of a new cause - that ofthe Natal colonists.

Though there were often strong differences of opinion betweenofficials, settlers and missionaries over what policies wereappropriate for dealing with the colony's now rapidly increas-ing African population, they were by and large united in thenotions that the safeguarding and expansion of their variousnewly acquired estates required the subordination of Africansto the political tutelage of Europeans. Like colonizing groupseverywhere, the Natal colonists sought to justify to them-selves and to others both their occupation of lands formerlyinhabited by other people, and their status as overlords orwould-be overlords of the indigenous peoples. The devastationstereotype was admirably suited to the version of Natal's his-tory which they developed to suit their ideological needs. Itjustified the presence of European colonists in Natal on thegrounds that they had established themselves in a largelyempty land. It justified their attempts - ultimately success-ful - to establish domination of the African population on thegrounds that the coming of white rule had put an end to theravages of the Zulu and inaugurated an era of peace andstability. It justified their continual demands for an in-crease in the size of the British garrison to defend themagainst the threat posed by the savage Zulu kingdom across theThukela.

Perhaps the clearest example of the early appropriation of thedevastation stereotype by settler ideologues is to be found inthe report, which appeared in 1853, of a major official com-

18. Kay, Travels and Researches, pp. 281, 341, 343, 344; D.Kotze, ed., Letters of the American Missionaries 1835-1838,Cape Town, 1950, p. 97; G. Champion, Journal of the ReverendGeorge Champion 1835-1839, ed. A. Booth, Cape Town, pp. 15,62; W. Boyce, Notes on South African Affairs, Cape Town, 19 71(reprint of 1st ed., Grahamstown, 1838), pp. x, 171, 173-4; B.Shaw, Memorials of South Africa, Cape Town, 19 70 (1st ed.London, 18 40), pp. 4 4-5; T. Arbousset & F. Daumas, Narrativeof an Exploratory Tour to the North-East of the Colony of theCape of Good Hope, Cape Town, 1968 (repr. of 1st English ed.,Cape Town, 1846), ch. 17, esp. p. 148.

mission appointed to enquire into the state of the colony'sAfrican inhabitants. The commission was dominated by repre-sentatives of the settlers: its declarations on theprecolonial history of Natal were of the kind that became com-mon in the literature over the next seventy years. After Shakahad become king, the commissioners reported, the Zulu became'a desolating scourge to all the surrounding tribes and na-tions within a circle of 500 miles'. He 'completely conqueredthe Natal tribes with immense slaughter, devastated the wholecountry and added it to his dominions, from the Itongati downto St. John's River'. The survivors were carried off and in-corporated in small groups into the Zulu people. 'The Nataltribes then ceased to have any separate national existence. . . • . (19)

Stated in these sentences is an important part of the histori-cal creed of an emerging settler society. The last sentence isperhaps the most significant. For intruding settlers, throwntogether from disparate backgrounds and attempting to estab-lish, first, a social existence and identity in territory pre-viously inhabited solely by African communities, and second,political domination over their Africa fellow-colonists, itwas important to minimize in their own eyes the rights to landof the people among whom they settled. The notion that the Af-rican societies which had formerly occupied Natal no longerhad any coherent existence was clearly convenient to theirpurpose.

The accounts of Natal's history which emerged from the pens ofcolonial-based writers in the 1850s and 1860s served to codifyviews such as those expressed by the commission. (20) Butthese writers did not simply reproduce the existingstereotype: they also placed it in a more elaborate historicalcontext. The Cape and British-based originators of thedevastation thesis had been concerned primarily to expand thegeographical orbit of the Cape's commerce, and had had littleinterest in investigating Natal's history beyond what wasnecessary for constructing the thesis in bare outline. Thesettlers, on the other hand, needed a more elaborate - anddenigratory - history of African societies in pre-European Na-tal: one which spelt out in unmistakeable terms that the his-tory of independent African societies was mostly one of warsand destruction. Civilization in Natal had begun with the com-

19. Proceedings and Report of the Commission Appointed to in-quire into the Past and Present State of the Kafirs in theDistrict of Natal.... Pietermaritzburg, 18 53, p. 6.

20. W.C Holden, History of the Colony of Natal. Cape Town,1963 (reprint of 1st ed., London, 1855), pp. 55-7; and ThePast and Future of the Kaffir Races. Cape Town, 19 6 3 (reprintof 1st ed., London, 1866), pp. 9-16, 20-28; J. Shooter, TheKafirs of Natal and the Zulu Country. London, 1857, ch. 8; L.Grout, Zulu-land; or. Life among the 2ulu-Kafirsr London, n.d(1863), chs. 7, 8.

8

ing of Europeans: for its impact to be the more fully appar-ent, the barbarism and savagery of 'Zulu' society had to berevealed in some detail. Hence in the works of settlerideologues the devastation thesis came to be situated in thecontext of a 'Zulu* history that extended back into the 18thcentury and forwards to the Boer victories in the late 1830sand the British annexation in the 1840s. (21)

The stereotype reinforced

The devastation stereotype as it had become entrenched insettler historiography in the 1850s and 1860s survived quiteunchanged into the early 20th century. In the late 19thcentury, as an alliance of British imperial interests andwhite settler interests in southern Africa set out to bringthe sub-continent's African societies once and for all underwhite domination, the stereotype began to penetrate more wide-ly. It moved beyond the accounts of colonial historians andlocal travellers to become established in reference worksranging from Natal schools text books to the EncyclopaediaBritannica, (22) and to become almost universally accepted inliterary discourse on southern Africa. (23)

Until this time the stereotype had remained based on the mostmeagre and tenuous empirical evidence, and set mainly in thecontext of the history of the Natal region. It existed verymuch as a formula reproduced by rote rather than as scientifi-cally argued history. In the late 19th and early 20thcenturies it was for the first time given powerful in-tellectual underpinning. The voluminous writings of GeorgeMcCall Theal rooted it firmly in the wider context of southernAfrican history, and the works of Father Alfred Bryant on'Zulu' history served to give it an apparently firm empiricalfoundation.

Except for his very earliest work, Theal's histories werestrongly pro-settler in their slant. (24) Like other settler

21. The first author to have developed a detailed account ofthis history seems to have been Shooter: see his Kafirs of Na-tal, pp. 249-60.

22. H. Bryan, Our Country: an Elementary History of Natal,London, n.d. (1909), pp. 28-31; Encyclopaedia Britannica. 9thed., vol. 24, Edinburgh, 1888, p. 828; and 11th ed., vol. 28,Cambridge, 1911, pp. 1051-52.

23. Only two authors in the period under review have beenfound in this study to have written in some respects directlycounter to the stereotype: A. Wilmot, The Story of the Expan-sion of Southern Africa. Cape Town, 1895, p. 126; J. Voigt,Fifty Years of the History of the Republic in South Africa(1795-1845). vol. 1 Cape Town 1969 (reprint of 1st ed.,London, 1899), pp. 185-92.

2 4. Recent critical studies of Theal are D. Schreuder, 'Theimperial historian as "colonial nationalist": George McCall

historians before and after him, he made no changes to the es-sentials of the stereotype, but an immensely important innova-tion on his part was to detach it from the history of the Na-tal region and set it in a nev context. Most previous writershad treated the pre-European history of Natal as part of theoverall history of that region: in Theal's works, by contrast,the upheavals in'Natal before and after 1820 were treated aspart of the history of the African societies of southern Afri-ca in the late 18th and early 19th centuries. In his mainwork, this history was covered in a block of chapters whichbroke into a sequence of chapters on the history of Europeansettlerdom and officialdom at the Cape. (25)

With hindsight it can be seen that this structuring estab-lished the preconditions for the emergence of mfecane-theory.By sharply segregating the histories of African and Europeansocieties in southern Africa, it allowed the impact of pre-18205 European settler and imperial influences on Africansocieties largely to be side-stepped, and the violence of the1820s to be attributed to purely 'African' causes. By linkingthe history of the Zulu kingdom in this period ethnically withthat of other African societies across southern Africa ratherthan regionally with that of neighbouring societies, both Af-rican and European, in eastern southern Africa, it allowed therise of Shaka and the Zulu kingdom to be portrayed as thespark which touched off a holocaust of intra-African violencewhich swept across the whole sub-continent. Where, in the ear-lier literature, the devastations supposedly caused by Shakahad usually been projected as having been confined to theeastern coastal regions, from Theal's time onward they wereuniversally seen as having affected much of the interior aswell. The devastation in Natal now came to be seen as simplyone aspect of a wider series of wars and migrations.

Schreuder has suggested that Theal's fairly novel sub-continent-wide perspective was rooted in the movement towardsthe creation of a common anti-imperial 'colonial' identitywhich was beginning to emerge among the English-speaking com-mercial classes and Afrikaner rural interests in the Cape inthe last two decades of the 19th century. In Theal's view, thecentral theme of southern African history was the march ofEuropean civilization, carried by Boer and British coloniststogether, across the sub-continent in the face of resistancefrom barbaric African tribes and of 'meddling' on the part ofthe British imperial government. As Schreuder puts it, 'It

Theal and the making of South African history', in G. Martel,ed., Studies in British Imperial History: Essays in Honour ofA.P. Thornton, London, 1986, pp. 95-158; C. Saunders, TheMaking of the South African Past: Maior Historians on Race andClass, Cape Town, 19 88, chs. 1-4.

25. G.M. Theal, History of South Africa since 1795. vol. 1,London, 19 08, chs. 14, 15.

10

was Theal, in fact, who historically invented white "South Af-rica"1, and his histories provided an important element in theemergence of an ideology of colonial nationalism and whitesupremacy. (26)

Theal's treatment of the history of African societies, it issuggested here, was a necessary counterpart of his concernwith the white colonial civilizing mission. If white colonistswere to be seen as the bearers of civilization, then, in apioneering colonial nationalist historiography that was emerg-ing when the process of bringing African people under the con-trol of white settlers was by no means over, it was importantto demonstrate in some detail that the culture of Africanpeoples was barbaric, and their history largely one of war anddestruction. The 'Zulu devastations' posited by Theal thusformed an important backdrop to what he saw as the essentiallycivilizing mission of the Great Trek.

If Theal provided the devastation stereotype with a new con-text, he added nothing to it by way of fresh evidence. Thefirst author to go much beyond the handful of primary accountson which the stereotype was based was Father A.T. Bryant,whose career as a writer of history overlapped with the latterpart of Theal's. Where Theal had been the first historian inSouth Africa to produce full-length studies based on in-tensive, if superficial and selective, archival research,Bryant was among the first writers to produce histories basedlargely on oral testimony collected from African informants.Between 1905 and 1929 he published a number of works on thehistory of the Natal-Zululand region which are still widelyused - usually quite uncritically - as primary sources. (27)These works were the first to set out detailed, though dis-continuous and confusing, accounts of the upheavals of thelate 1810s and early 1820s, but, far from challenging thestereotype, Bryant presented his information in such a way asto reinforce it.

26. Schreuder, 'The imperial historian as "colonial nation-alist"', in Martel, ed., Studies in British Imperial History,pp. 96-8, 114-15, 129-33, 138-47. The quotation is from p.97.

27. Bryant's first major historical work was the essayentitled 'A sketch of the origin and early history of the Zulupeople', which was published as a preface to his Zulu-EnglishDictionary, Mariannhill, 1905, pp. 12*-66*. This was followedby a series of articles published in the newspaper IzindabaZabantu in 1910-13; these were reprinted after the author'sdeath in A.T. Bryant, A History of the Zulu and NeighbouringTribes, Cape Town, 1964. His widely influential Olden Timesin Zululand and Natal, was published in London in 1929. Hismajor ethnographic study, The Zulu People as They Were beforethe White Man Came, Pietermaritzburg, 1929, begins with twohistorical chapters.

11

Bryant•s own work indicated that several 'waves' of non-Zuluinvaders had swept through much of Natal some time before theadvent of Shaka's Zulu armies, and that Zulu incursions had infact been confined to a comparatively small area. But, inkeeping with the established stereotype, he all but submergedthe role played by these precursors in favour of an emphasison the doings of the Zulu. His dramatized and often hyperboli-cal description of Shaka's wars and conquests, one inheriteddirectly from the settler stereotype, underscored the ideathat Shaka and his 'Zulu murderers' were the main destroyersof Natal. (28) In addition, Bryant's method of presenting theregion's history through a recital of the individual historiesof the numerous chiefdoms of Natal had the effect of repeated-ly bringing Shaka and the Zulu into the narrative in the roleof conquerors, exterminators, and tribute-takers. For thereader trying to pick his or her way through the mass of oftencontradictory historical detail which Bryant presented, theZulu could hardly have emerged as anything other than the mainvillains of the piece. In spite of the empirical richness ofhis account, he was unable to break with the establishedstereotype.

The stereotype sanctified

In the works of the liberal academic historians who dominatedthe writing of South African history in English from the 1920sonward, interest both in regional history and in the historyof autonomous African societies began to dry up. By this timethe affairs of the recently established Union of South Africawere coming to overshadow those of its constituent parts inpolitical significance, and the social and political impact ofthe country1s ongoing industrial revolution was preoccupyingits intellectuals as well as its politicians. The consequencewas another new turn in the history of the devastationstereotype.

The emergence of liberal historiography in the post-World WarI era needs to be seen against the background of the erosionof the bases of the old agrarian-commercial order, the spreadof poverty in the African reserves, the large-scale migrationof Africans to the urban areas and the resultant problems ofsocial control, and the formation of potentially formidable,if unstable, political alliances between sections of the ex-ploited African proletariat and frustrated and militant ele-ments of the emerging African petty bourgeoisie. For liberalintellectuals these developments posed a profound dilemma.(29)

28. The phrase 'Zulu murderers' occurs in his Dictionary, p.49*.

29. On the nature of South African liberalism between the warssee P.Rich, White Power and the Liberal Conscience: RacialSegregation and South African Liberalism. Johannesburg, 19 84,chs. 1-3; S. Dubow,'"Understanding the native mind": the im-pact of anthropological thought on segregationist discourse inSouth Africa, 1919-1933', unpublished conference paper, Uni-

12

On the one hand they were concerned about the possible politi-cal consequences of increased legal entrenchment of dis-crimination against Africans, and of the increasing suppres-sion of the material and political aspirations of the African•elite1. On the other, they were fearful about the threatwhich, to their minds, the processes of African urbanizationand proletarianization presented to 'civilized values' (read•capitalist order1) in South Africa. The response of liberalhistorians was two-fold. (30) In reaction against the racismof settler historiography they turned their attention to thehistory of 'race relations' in South Africa, and, to enableliberals the better to grapple with the complex ramificationsof the 'native problem1, they began to focus on the nature ofthe historical forces which, as they saw it, had drawn Afri-cans and whites into a common society. Those fields of his-tory which offered little or nothing by way of an explanationof these issues were neglected.

As a result, the writing of precolonial history, whether lo-cally focussed as in Natal colonial historiography, or broadlyfocussed as in Theal's works, virtually died out. In turn thismeant that the devastation stereotype escaped the criticalscrutiny of the first generation of academic historians toemerge in South Africa. With their main fields of interestlying elsewhere, these historians were content to absorbTheal's generalized view of the 'Zulu devastations' into theirown work, even if in the process they were concerned to stripit of its more racist forms of expression. Thus in the worksof writers like Walker, Agar-Hamilton, Macmlllan, Hattersley,De Kock, De Klewiet and others, the notion remained quite un-challenged that Natal had been swept almost clear of in-habitants during wars of extermination waged by Shaka and the

versity of the Witwatersrand, 1984', pp. 1-22; S. Dubow,'Race, civilisation and culture: the elaboration ofsegregationist discourse in the inter-war years', in S. Marks& S. Trapido, eds., The Politics of Race, Class and Nation-alism in Twentieth-Century South Africa. London, 19 87, ch. 2.Less critical studies are R. Elphick, 'Mission Christianityand interwar liberalism', in Butler et al., eds.. DemocraticLiberalism in South Africa: Its History and Prospect, Mld-dletown. Conn., & Cape Town, 1987, ch. 3; and J. Butler, 'In-terwar liberalism and local activism', in Butler et al., eds..Democratic Liberalism, ch. 4.

30. On liberal historiography from the 1920s to the 1950s seeM. Legassick, 'The frontier tradition In South African histor-iography', in S. Marks & A. Atmore, eds., Economy and Societyin Pre-Industrial South Africa. London, 1980, ch. 2; C.Saunders, 'Liberal historiography before 1945', in J. Butleret al., eds., Democratic Liberalism In South Africa: Its His-tory and Prospect. Middletown, Conn., & Cape Town, 1987, ch.7; J. Butler & D. Schreuder, 'Liberal historiography since1945', in Butler at al., eds.. Democratic Liberalismf pp. 148-56; Saunders, Making of the South African Past, chs. 5-9.

13

Zulu. (31)

But the stereotype did not survive in early liberal histori-ography simply by default. Part of the neglect of Africanhistory by liberal historians from the 1920s onward must alsobe put down to their often elitist attitude to the culture ofthe African underclasses, and to their ambivalence on the is-sue of what political rights to accord to Africans. (32) Evenif liberals rejected the notions put forward by racialtheorists in the late 19th and early 20th centuries to the ef-fect that the African 'race' was incapable pf 'catching up'with the civilization of the European race, until at least the1950s many of them still thought of cultural differences instrongly evolutionist and hierarchical terms. Africans wereexpected to 'adapt' their own cultures in the direction ofEuropean culture, and, in effect, to reject their own past,which had little, if anything, to do with 'progress'. Fromthe 1920s, African history was largely discounted by liberalintellectuals, and the study of African societies was in-creasingly seen as belonging to the emerging discipline ofanthropology rather than to history. Now that autonomous Af-rican societies had been thoroughly subjugated, their versionsof history did not have to be contested, as had been necessaryin the era of white settlement: their overlords could simplydisregard their history altogether.

In the years between the two world wars one of the centralconcerns of South African liberal intellectuals had been totry to identify the 'problems' in white-black 'race relations'which were seen as preventing orderly social evolution towardsthe 'civilized' society which they ardently hoped for. Liber-al historians had been concerned to look for the historicalroots of those problems as one means of working towards theirsolution. After the National Party's electoral victory of1948, and the further entrenchment of racial discriminationand repression, the problem for liberal historians came to be

31. E. Walker, A Modern History for South Africans, Cape Town,1926, p. 225; E. Walker, A History of South Africa, London,1928, pp. 182-3; J. Agar-Hamilton, The Native Policy of theVoortrekkers, Cape Town, n.d. (1928), pp. 2-3; W.M. MacmilIan,Bantu. Boer and Briton: the Making of the South African NativeProblem. London, 1929, pp. 13-14; A.F. Hattersley, South Afri-ca 1652-1933. London, 1933, p. 75; M.H. de Kock, The EconomicDevelopmentof South Africa, London, 1941, p. 50; C.W. deKiewiet, A History of South Africa Social and Economic. Ox-ford, 1941, pp. 49-50.

32. On this ambivalence see Rich, White Power, passim;Saunders, 'Liberal historiography before 1945', in Butler etal. eds., Democratic Liberalismf pp. 139, 145; Butler &Schreuder, 'Liberal historiography since 1945', in Butler etal, eds., Democratic Liberalism, pp. 154-6; D. Irvine, 'TheLiberal Party, 1953-19681, in Butler et al., eds., DemocraticLiberalism, pp. 117-19, 12 5-30.

14

why whites had taken the 'wrong road1 and refused to sharepower with blacks. The focus of their work swung furthertowards political and constitutional history, with a con-comitant further decline in the attention paid to African his-tory. (33)

If early liberal historiography showed little interest eitherin the history of African societies, or in the history of Na-tal before the advent of Europeans, these subjects receivedeven less attention in the Afrikaner nationalist histori-ography that was emerging from the early years of the 20thcentury onwards. Whether it was in popular histories or in theworks of the professional historians who were becoming estab-lished in the Afrikaans-language universities after World WarI, the focus of this historiography was overwhelmingly on thehistory of Dutch-Afrikaner societies from the Great Trek tothe South African war of 1899-1902. So far as it touched onthe history of pre-Trek Natal, it simply reproduced thestereotypes entrenched by Natal settler historiography and byTheal and other contemporary writers, though now placing themin the context of Voortrekker history rather than in the con-text of Natal's history or the history of African societies.In this context, the devastation stereotype served convenient-ly to portray Natal as having been emptied of population be-fore the coming of the Voortrekkers. (34) Together with con-temporary liberal historians, then, Afrikaner nationalist his-torians carried the devastation stereotype, set in the contextof settler history, from the early 20th century through intothe 1960s.

The stereotype africanized

It was not until the 1960s that historians turned back towriting the history of the African societies of southern Afri-ca. As is well known, its revival was an aspect of thegrowth of interest in African history in general which ac-companied the political decolonization of most of the con-tinent north of the Zambezi. The ending of colonial rulestimulated an eager demand among African political activistsand intellectuals, and among sympathizers in Europe and NorthAmerica, for a 'decolonized' African history, one which wouldrescue Africans from the virtual oblivion to which they had

33. Saunders, 'Liberal historiography before 1945', in Butleret al. , eds., Democratic Liberalism, p. 147; Butler &Schreuder, 'Liberal historiography since 1945', in Butler etal . , eds., Democratic Liberalism,, pp. 151-4, 160-1.

34. S. Gie, Geskiedenis vir Suid-Afrikar vol. 2, Stellenbosch,1928, p. 306; E. Jansen, Die Voortrekkers in Natal: Qpstelle,Cape Town, 1938, p. 1; A. du Plessis, 'Die republiek Natalia',Archives Year Book for South African History. 1942, part I,Cape Town, 1943, ch. 1; A. van der Walt, 'Die Groot trek tot18381, in A. van der Walt et al., eds., Geskiedenis van Suid-Afrikar vol. 1, Cape Town, 1951, p. 264.

15

been consigned by colonial historiography, and which would em-phasize African 'achievements*. An alliance of African na-tionalist and metropolitan liberal historians moved to meetthis demand by producing history which took as its mainthemes the emergence of great states in precolonial Africa,the mounting of resistance to European conquest and colonialrule, the growth of African nationalism, and the role of Afri-can elites.

Within this context there was published in 1966 the first worksince Theal's to attempt a broad synthesis of the history ofthe upheavals which had taken place in south-east Africa inthe 1820s and 1830s. This was J.D. Omer-Cooper's full-lengthstudy, The Zulu Aftermath: a Nineteenth-Centurv Revolution inBantu Africa. (35) In important respects Omer-Cooper's ac-count was similar to Theal's. It saw the violence of the peri-od as having emanated from a single epicentre, the Zulu king-dom, and as having radiated outward across much of southernand central Africa. Omer-Cooper attributed the ultimate causesof the upheavals to a build-up of population pressure insouth-east Africa in the later 18th century, rather than tothe personality of Shaka, as Theal had done, and he wentbeyond Theal in seeing the effects of the violence as havingextended over wide areas of Central Africa, and as having per-sisted into the latter half the 19th century. But, as thetitle of his book indicates, like Theal he was looking for acompendium explanation to cover what he saw as a single his-torical phenomenon. (36)

In two respects, though, Omer-Cooper introduced major con-ceptual innovations into the treatment of the subject. In thefirst place, in sharp contrast to Theal, who had emphasizedthe violence and bloodshed that had accompanied the upheavalsin order to portray them as an indication of African barbarismand savagery, Omer-Cooper depicted them in positive terms as'one of the great formative events of African history', as anepisode of 'nation-building' on the part of 'a galaxy of greatleaders'. (37) In the second place, he gave the upheavals asingle label, one which has stuck both in academic and in pop-ular usage ever since. This was the term 'mfecane1, which hegave as meaning 'the wars and disturbances which accompaniedthe rise of the Zulu'. (38) The word had been used sporadical-ly in the literature since the 1920s, though without a clearlydefined meaning. (39) Omer-Cooper both standardized its mean-ing and projected it into general usage.

35. London, 1966.

36. Omer-Cooper, Zulu Aftermath, introduction & pp. 19-27.

37. Ibid., pp. 4-7.

38. Ibid., p. 5n.

39. Cobbing, 'The case against the mfecane', unpublishedpaper, University of the Witwatersrand, p. 5.

16

In his treatment of the region south of the Thukela, to whichhe devoted a chapter, Omer-Cooper drew heavily on Bryant'sOlden Times in Zululand and Natal. (40) There was nothing inhis account that explicitly contradicted the devastationstereotype; if anything. The Zulu Aftermath served to give itfurther academic respectability, and, by presenting it as anintegral part of the long.-established and now revamped notionfreshly packaged as the 'mfecane1, to publicize it more widelythan ever. From this time on, the history of the stereotypewas closely intertwined with that of mfecane-theory.

Few works on southern African history have had so immediateand widespread an effect as The Zulu Aftermath. As Cobbing haspointed out, within a few years of the book's publication boththe term mfecane (or difaqane) and the notion that the mfecanewas one of the central events of southern African history hadbecome embedded in Africanist discourse outside South Africa.(41) It was widely established in general histories of Africa,in academic articles and monographs, in university and schoolstext books, and once again, though now in a quite differentcontext, in the Encyclopaedia Britannica. (42)

In the late 1960s and early 1970s South African academics be-gan to catch on to the notion of the mfecane. The first to doso were Afrikaner nationalist historians, who were quick tospot the support which mfecane-theory lent to the ideological-ly important notion that the first white settlers in the inte-rior of southern Africa had moved into a land largely depopu-lated by intra-African warfare. (43) A little later, under theinfluence of overseas Africanists, liberal writers began toincorporate the mfecane into their work as the fons et oriooof the processes of African 'state-formation' (another termfor 'nation-building') round which they wrote the history ofAfrican societies in south-eastern Africa in the first half of

40. Omer-Cooper, Zulu Aftermath, ch. 10.

41. Cobbing, 'The case against the mfecane', unpublishedpaper. University of the Witwatersrand, pp. 5-7; and 'The mythof the mfecane', unpublished paper, pp. 8-9.

42. S.M. (Shula Marks), article on the history of southern Af-rica in Encyclopaedia Britannica. vol. 17, Chicago, 1974, p.281.

43. C. Muller, 'The period of the Great Trek, 1834-1854', inC. Muller, ed., Five Hundred Years: a History of South Africa,Pretoria & Cape Town, 1969, p. 125; D. Ziervogel, fThe nativesof South Africa', in Muller, ed., Five Hundred Years, pp. 445-8; F. van Jaarsveld, Van Van Rlebeeck tot Verwoerd 1652-1966,Johannesburg, 1971, ch. 7; J. Bruwer, article on Shaka inStandard Encyclopaedia of Southern Africa, vol. 9, Cape Town,1973, p. 598; C. Muller, Die Oorsprona van die Groot Trek,Cape Town, 197 4, esp. pp. 7 4-83.

17

the 19th century. (44) They were followed by some African na-tionalist writers, particularly those sympathetic to Zulu eth-nic nationalism, for whom the 'nation-building' aspects of themfecane were an obvious attraction. (45)

By the later 1970s the mfecane was an established 'fact' ofsouthern African history. Though by then a reaction againstthe more uncritical assumptions and assertions of Africanisthistory was manifesting itself among liberal and radical-revisionist historians alike, the mfecane lived on in SouthAfrica and abroad virtually unchallenged. In 1983 mfecane-theory found standardized form in the first academic diction-ary of south African history to be published. (46) In 1986 itachieved coffee-table status, (47) and in 1987, some twentyyears on from The Zulu Aftermath, it was re-invigorated byOmer-Cooper himself in a new text book on South African his-tory. (48) In the late 1980s liberals, radicals, African na-tionalists, and Afrikaner nationalists continued in an unlike-ly, if unwitting, alliance, some propounding, some merely ac-cepting, but virtually none challenging the validity of thenotion of the mfecane.

Mfecane-theory emerged at a time when liberal and African na-tionalist historians outside southern Africa were seeking tobreak away from racist and patronizing colonial cliches aboutAfrican culture and African history. It is easy to understandwhy the notion of the mfecane as a period of African 'nation-building* caught on so rapidly among them, and why it survivestoday in the uncritically Africanist histories that continueto be produced. It is easy, too, to understand the continuingattraction of mfecane-theory's 'depopulation' thesis for theideologues of apartheid.

More problematic is the failure of contemporary critical lib-

44. For example the present author's Bushman Raiders of theDrakensbera 1840-1870, Pietermaritzburg, 1971, pp. 15-17;several of the articles in C. Saunders & R. Derricourt, eds.,Beyond the Cape Frontier, London, 1974; C. Webb, fOforthodoxy, heresy and the difaqane', unpublished paper, Uni-versity of the Witwatersrand, 1974; and 'The Mfecane1, in Per-spectives on the Southern African Past. Centre for AfricanStudies, University of Cape Town, 1979, ch. 9; T. Davenport,South Africa: a Modern History, Johannesburg, 1977, pp. 10-17.

45. J. Ngubane, 'Shaka's social, political and militaryideas', in D. Burness, ed., Shaka King of the Zulus in AfricanLiterature. Washington, D.C., 1976, pp. 140, 147.

46. C. Saunders, Historical Dictionary of South Africa,Metuchen, N.J., & London, 1983, pp. 107-8.

47. Edgecombe, 'The Mfecane or Difaqane1, in Cameron & Spies,eds., Illustrated History, ch. 9.

48. J.D. Omer-Cooper, History of Southern Africa, London,1987, ch. 4.

18

eral and radical-revisionist scholarship to challenge mfecane-theory. At a superficial level this failure can be explainedin terms of the general decline of interest among scholarsabroad in African history since about the mid-1970s. (49) Re-searchers are thinner on the ground than they used to be inthe days of the Africanist boom, and they have little incen-tive to tamper with what appears to be a coherent and veil-grounded set of notions which puts the precolonial Africanstates of southern Africa firmly on the historical map. InSouth Africa itself, after a brief flowering in the 1970s, in-terest in the region's precolonial history has waned as liber-al and radical historians have increasingly focussed their re-search and debates on the effects of capitalist penetration insouthern Africa from the late 19th century onward. Partly,then, mfecane-theory survives today by default.

But, at a deeper level of explanation, it survives among lib-eral historians, as Cobbing has argued, because it functionsto obscure the processes by which a white property-owningclass came to be politically dominant and in possession ofmost of the land south of the Limpopo river. (50) By omittingthe role of white agency in the upheavals of the 1820s and1830s, and by attributing them ultimately to the rise of theZulu kingdom, mfecane-theory is able to portray them as a con-sequence of internecine African conflict. African agency thusbecomes responsible for opening the way for the penetration ofwhite settlers into a largely 'empty' interior, and the land-grabbing of whites later in the century can be dovn-played.Cobbing's argument perhaps overstresses the cohesion andstrength of white settler societies in the 19th century, butits central point seems essentially correct. For liberal de-fenders of South Africa's capitalist order, as well as for theideologues of South Africa's bantustan policies, mfecane-theory provides an ideologically incontrovertible explanationof the historical basis of South Africa's present-day patternsof land distribution.

Among radical historians too mfecane-theory survives todaypartly for ideological reasons. The 'structuralist • approachwhich was dominant among radical writers in South Africa inthe 1970s did not encourage detailed scrutiny of historicalevidence. While the reaction on the part of many contemporaryradical historians against the often reductionist analysesthat were generated by their predecessors has made for theproduction of a far more textured and nuanced kind of history,

49. T. Ranger, 'Towards a usable African past', in C. Fyfe,ed., African Studies since 1945. London, 1976, p. 17; P.Curt in, 'African history', in M. Kammen, ed., The Past beforeUs: Contemporary Historical Writing in the United States,Ithaca & London, 1980, p. 115.

50. Cobbing, 'The case against the mfecane1, unpublishedpaper. University of the Witvatersrand, pp. 1, 7-8, 16, and'The myth of the mfecane', unpublished paper, pp. 1, 9-10, 30

19

it has also meant a loss of much of the political punch whichradical history carried a decade ago. With their focus oftenon microstudi.es, and with their tendency to be suspicious ofschematizing and generalization, present-day radical histori-ans are often less overtly concerned than the previous genera-tion was to identify and hammer away at the ideological props,such as mfecane-theory, which help sustain the current racialand social order in South Africa.

Conclusion

A hundred and sixty years after it first surfaced, thedevastation stereotype lives on, embedded now in mfecane-theory. Cape merchant interests created it in the 1820s and1830s on the basis of hearsay evidence. Natal settlers fromthe 1840s onward, the first South Africanists in the late 19thand early 20th centuries, and Afrikaner nationalist historiansfrom the early 20th century onward all had a common vested in-terest in keeping it alive. Early liberal historians in SouthAfrica, with their attentions elsewhere, incorporated it as anelement in what they saw as the not very important history ofthe country's African underclasses. Later liberal historians,first outside and then inside South Africa, helped resuscitateand reconstruct it in a way that would be aceptable to emerg-ing African nationalist elites. After a brief period of con-cern with developing new approaches both to pre-industrialhistory and to the macrohistory of southern Africa, the major-ity of radical historians turned away towards the more recentpast and towards a partially depoliticized social history,leaving the stereotype intact. Present-day writers of allshades of opinion continue to pick it up from the previousliterature and to incorporate it into their own work withoutattempting to seek empirical verification for it. Natal'smfecane exists today by virtue not of historical argumentationbut of uncritical repetition of a racist and elitist myth.